AI decision making risks explained: Lessons on AI accountability and governance: Why did IBM warn in 1979 that computers should never make management decisions and why is it relevant now as AI takes over?
How Does IBM’s 1979 Liability Warning Still Apply Today?
One of the key points of the IBM Training Manual was, according to IBM, “A computer can never be held responsible, so a computer should never make management decisions.”
AI Shifts from Support Tool to Decision Making in 2026
In 2026, AI is no longer limited to analysis or automation alone. It is now integrated into enterprise workflows, supporting forecasting, customer operations, risk assessment and internal decision systems.
Recent industry research shows that AI adoption has reached major levels across organizations, with most companies already using or scaling AI in some form across business functions such as customer service and decision support, according to the ArXiv report.
In practical terms, artificial intelligence no longer only suggests actions, but increasingly shapes decisions itself.
The Rise of Autonomous “Agent-Like” Artificial Intelligence Systems
Meanwhile, new enterprise trends show AI systems evolving into more autonomous “agent-like” tools that can execute workflows, increasing both efficiency and governance challenges, according to a TechRadar report.
Governance Gap: AI Adoption vs Oversight
As AI adoption accelerates, governance and oversight are struggling to keep up. Recent reports highlight that organizations are moving faster on deployment than control frameworks, creating gaps in monitoring, accountability and risk management.
This is especially important as businesses scale AI for core business decisions rather than just support functions.
Industry researchers now describe this as a structural change; Companies have powerful models and tools, but they still lack consistent systems to ensure accountability when AI influences real decisions, according to the ArXiv report.
Who is Responsible When AI Gets It Wrong?
As AI becomes more involved in management decisions, determining responsibility becomes more difficult.
If an AI-driven recommendation leads to financial loss, operational error, or flawed business selection, liability could potentially spread to:
Managers who approve the use of artificial intelligence
Engineers who built the system
Teams that deployed this
Or the organization as a whole
According to the ArXiv report, researchers describe this as a growing governance and trust gap where organizations struggle to prove who is responsible for AI-driven outcomes.
Why Humans Are Still in Control: AI Lacks Ethical and Contextual Judgment
Despite rapid adoption, most organizations are not completely delegating decision authority to AI.
AI systems are widely viewed as strong at processing data, detecting patterns, and increasing speed, but are seen as weaker when it comes to ethical reasoning and context-sensitive decisions.
That’s why many enterprise systems still rely on human oversight; Here, AI generates insights but humans verify final decisions before implementing them.
Transitioning from AI Adoption to AI Governance and Control
As AI becomes more involved in decision-making processes, companies are now shifting their focus from “AI adoption” to “AI control.”
New governance frameworks are being developed to ensure transparency, auditability and traceability across AI systems, especially as they become autonomous, according to a report from IBM.
According to the ArXiv report, some of the latest research even suggests continuous monitoring systems that monitor AI behavior in real time to ensure accountability throughout its lifecycle.
In parallel, companies are increasingly investing in governance platforms designed to manage compliance, risk and oversight as AI scales across operations, according to the TOI report.
FAQ
What is the IBM 1979 warning about?
He says computers should not make management decisions because they cannot be held accountable.
Is AI really making decisions in companies today?
Artificial intelligence is increasingly influencing and shaping decisions in businesses.
